Image: Recycling logo by gustavorezende, released to public domain
Warning: long post. There’s a TL;DR in the Summary at the end.
Is recycling Methods text from an old paper, to use in a new paper that applies the same techniques, efficient writing – or self-plagiarism?
We’ve all had the dilemma. You write two papers that use (at least some of) the same methods. For the first paper, you craft a lovely, succinct, clear explanation of those methods. For the second paper, you’d like to just cut-and-paste the Methods text from the first one. Can you? And should you?
I’m sorry, but I think the answer is usually “no”*.
It’s an old, old question, but it still gets asked. A quick googling (try “self plagiarism methods”) will find you lots of material to read. The problem is, I’ve seen few answers that seem convincingly complete. Even more unhelpfully, many seem to be written by folks who have very strong opinions in one direction or the other. Often, a promising answer will turn into an indignant insistence that of course you can reuse your own words, and that the term “self-plagiarism” is nonsensical because plagiarism is intellectual theft, and you can’t steal from yourself. I certainly understand where this indignation is coming from, but I don’t think it holds up to careful examination. We’re going to try some of that (careful examination, not indignation, I mean), below.
I think there are three major pieces to any careful argument about reusing Methods text. You have to consider all three before coming to a decision about whether reusing Methods text is a good idea. Roughly, #1 and #2 determine whether you can reuse text, while #3 is about whether you should. Here they are:
This is the most obvious point and the one most online arguments turn on. Important note: I am not a copyright lawyer, and none of this constitutes legal advice.
For many published papers, copyright is not held by the authors; rather, it’s assigned to the journal (or its publisher)**. If your paper falls into this category, then no, you can’t reuse “your own” words in another paper; in a sense, they aren’t “your own” words any more. Reusing text that’s under someone else’s copyright is surely plagiarism, even if you are the original writer of that text!
Of course, for other published papers the authors retain copyright. If your paper is one of these, can you reuse the Methods text? Sometimes, to be sure. But you might have retained copyright but executed a license with the journal that would be violated by reusing text. More interestingly, what if you had co-authors? In the U.S., at least, coauthors would be joint holders of copyright, and it seems that one coauthor can probably reuse the material as long as any commercial proceeds (ha!) are shared. Whether this is true worldwide, I don’t know. And what if that reuse involves assigning copyright to the publisher of the second paper? Would all coauthors of the first paper have to assent? I don’t know that either, but the important point is probably just that the legal weeds have the potential to grow thick.
2. Journal policies
Having the legal right to reuse text isn’t the end of the story. Journal policies may preclude reuse of Methods text even when the author has clear control over the copyright. For example, here’s a statement from the journal Applications in Plant Sciences, which prohibits reuse of text (outside some narrow and limited circumstances) even if such reuse is legal. Now, you might well insist that if you own the copyright, you’re legally entitled to reuse your text – but even if that’s true, you aren’t legally entitled to have your paper accepted by Applications in Plant Sciences (or by any other journal holding a similar policy). Journals are free to hold any number of editorial policies (they can prohibit use of the letter “w” if they like); you are equally free not to submit to them.
While it isn’t a journal policy, the guidelines on text recycling from the Committee on Publication Ethics (COPE) are worth reading. These guidelines are essentially advice offered to journals. Here’s the core of COPE’s advice:
“Use of similar or identical phrases in methods sections where there are limited ways to describe a method is not unusual; in fact text recycling may be unavoidable when using a technique that the author has described before and it may actually be of value when a technique that is common to a number of papers is described. Editors should use their discretion and knowledge of the field when deciding how much text overlap is acceptable in the methods section.”
“The reuse of a few sentences is clearly different to the verbatim reuse of several paragraphs of text…” (COPE 2016)
This is more permissive than some journal policies, but how much is not terribly clear cut. It defers to editorial discretion on “how much text overlap is acceptable” (and suggests that this will vary across manuscript sections and across fields). One thing does seem clear: it’s not a license to reuse long blocks of text – notice in the first quotation reference to “similar or identical phrases” (emphasis mine), and in the second to “a few sentences”.
Bottom line: journal policies will vary, but may well be more restrictive than copyright considerations alone. (Even if not more restrictive, of course, neither COPE’s advice nor a journal policy can trump copyright.)
3. Readers’ needs
Even if both copyright law and journal policy will allow you to reuse Methods text, should you? Few pieces about text reuse consider the reader’s perspective on the issue. And the reader’s perspective, I think, provides another reason not to reuse text even when you can. Sure, two papers may use the same study system, or some of the same methods. But unless you’re salami-slicing rather finely, it seems unlikely that the reader needs exactly the same description of exactly the same details with exactly the same emphasis in exactly the same order. It’s easier to write that way, sure; but so what? Each paper should tell its own story to its own readers, and I think that will usually mean custom rewriting even of shared methods.
As an example: imagine a bird mistnetting study in which the first paper reports species densities and diversity, while the second reports on post-capture behaviour and blood cortisol levels. Details about the duration and conditions under which birds were held after capture but before processing wouldn’t help the reader of the first paper, but would be critical for readers of the second. The same mistnetting Methods repeated would serve neither reader well. (There may, of course, be exceptions to this. I suspect they will mostly involve completely standardized methods like PCR amplification. I wonder how often these will be longer than a paragraph or so, and whether they couldn’t be easily addressed in other ways than verbatim repetition.)
So if you can’t reuse text, what do you do instead?
Fortunately, it isn’t very difficult to avoid reusing the same Methods text. There are at least five options:
- English is a rich language, and there’s never just one good way to say anything***. I’ve written the same description of my goldenrods-and-gallmakers study system at least 8 times and I’m not running out of ways yet. Of the five options, this is my default and I think it’s usually the best – especially because it dovetails so nicely with the reader’s need for different details and different emphasis each time. UPDATE: following some pushback against this option, I expand on it here.
- Cite your first paper in your second. If the methods really were identical, offer a “Briefly we…” followed by “For full details, see Lee et al. 1982”. If they weren’t identical, try “We did X following methods in Lee et al. 1982, except that…”. Note that serves readers best if (1) they don’t need every detail to understand the current paper; (2) the cited paper is easy to find; and (3) the cited paper doesn’t pull the same trick, starting a long chain of citation-tracking back through the literature.
- Cite a standard methods paper. Usually this will involve “methods as in… except”, and it’s subject to the same qualifications as when you cite your own older paper.
- Deposit the full Methods detail elsewhere, and cite that with each new paper. That’s a role that can be played by protocols.io, for example, although there are other options. This neatly avoids the chain-of-citations problem. (Note that this practice may or may not pass muster with the publishing journal.)
- Quote the older paper directly. We don’t use direct quotation much in the sciences, and I’m not sure I’ve ever seen this done, but in principle it ought to work. Directly quoting the Methods from the older paper (with appropriate citation, of course) should overcome the copyright problem, so long as the quoted material is of a quantity that makes it fair dealing (that’s the Canadian term; it differs in other jurisdictions). Would science journals go for this? I don’t know.
I’m afraid this got pretty long, so here’s the TL;DR. Often, copyright considerations mean you can’t reuse Methods text from previous papers. Even when it’s copyright-appropriate, journal policy may not allow it. And even when copyright and journal permit, reusing text may not serve your readers well. I know that means a little (but only a little) extra work in writing. Sorry about that.
What angles on this have I missed? Fill us in, please, in the Replies.
© Stephen Heard April 3, 2018
This post is based on, but significantly expanded from, my treatment of the issue in The Scientist’s Guide to Writing (my guidebook for scientific writers).
There’s a lot of material on self-plagiarism out there. Here are some other links – some complementary, some as counterpoint. Please leave further suggestions in the Replies and I’ll add relevant ones to this list.
*^Experience suggests that this will make some people very angry. Many of those people will have already stopped reading, having their minds made up and being uninterested in thinking about the issue further. At least one of them will make some incandescent comments, probably on Twitter, about how stupid this post is. If I’m lucky, they’ll do it as a subtweet and I won’t see it. (To be clear: thoughtful disagreement is both possible and welcome! It’s the other kind of disagreement this footnote is about…) UPDATE: it took less than 90 minutes for this to happen. People are sometimes quite predictable!
**^Note that whether or not it should be this way is a totally different argument, and one that I’m not tackling today. As an interesting complication, in some cases copyright never belongs to the author to assign. In Canada, papers by employees of the federal government are always copyright “Her Majesty the Queen in right of Canada” (more or less, copyright belongs to the government, not the author, likely because the paper is considered a work for hire). This raises the somewhat delicious possibility that in a later publication one could be plagiarizing the Queen.
***^Writers who don’t understand this sometimes find themselves struggling with writer’s block, because they believe there’s a single best way to say Thing X, and they haven’t found it, and they won’t go on until they do.